Construction and expression of a reshaped VH domain against human CD28 molecules

Compared with the amino acid sequence of a mouse anti-human CD28 VH domain antibody, the two most homologous sequences of human antibodies were pulled out from Genbank. One of them was used as the main template for the framework regions of the reshaped VH domain. While the original mouse antibody CDRs were inserted into the human acceptor FRs, some residues in human acceptor FRs, which were different from those of the original mouse FRs in corresponding positions, were then determined or, alternatively, mutagenized to their conservative properties in kappa classification. Based on the amino acid sequences of the designed VH domain, the nucleotide sequence was deduced by using E. coli bias codons. The sequence was split into ten 30 to 60 nucleotide fragments for synthesizing, then annealed and amplified by overlap PCR. Taq DNA polymerase was used in a buffer with high Mg2+ concentration to induce more random mutations, both in FRs and CDRs. A phage display library was constructed by cloning these PCR products. After three rounds of panning, several reshaped VH with high antigen binding activity were obtained. One of them had the same CDR amino acid sequences as that of the original mouse VH domain. Further study showed that it retained a high antigen binding affinity after being expressed in E. coli BL21 (DE3).